Alcoholic liver disease eventually causes liver failure. The only curative therapy for patients with liver failure due to alcoholic liver disease is liver transplantation. Liver cell therapy is not effective in these patients because alcoholic liver disease is invariably associated with liver fibrosis, which impairs the engraftment of transplanted cells. Because of the long-standing shortage of donor livers, many patients with alcoholic liver disease die while waiting for liver transplantation. To improve the outcomes of these patients, we propose to work around liver fibrosis as a barrier to liver cell therapy. We hypothesize that the function of fibrotic livers can be improved by in vivo conversion of the cells that make up the fibrotic tissue into cells with hepatocyte function. Our hypothesis rests on the recent identification of hepatic transcription factors that can convert fibroblasts into therapeutically effective hepatocyte-like cells (iHeps) and success with in vivo conversion of fibrotic heart tissu into functional parenchyma. To establish in vivo iHep generation as a therapy for liver fibrosis, we will pursue 2 aims: (1) To convert myofibroblasts into iHeps in vivo. We will deliver the hepatic transcription factors to myofibroblasts with adenoviral vectors, which have been shown to efficiently transduce myofibroblasts in rat and mouse models of liver fibrosis. Most myofibroblasts are quiescent in advanced liver fibrosis, which will prevent loss of nonintegrating adenoviral vectors before conversion of myofibroblasts into iHeps is completed. Because myofibroblasts are the main source of collagen in the liver, our strategy is expected to not only increase the number of cells with hepatocyte function, but also prevent further collagen deposition in the liver. (2) To establish the use of nonintegrating adenoassociated viral (AAV) vectors for conversion of myofibroblasts into iHeps in vivo. Conventional adenoviral vectors are highly immunogenic and therefore not safe for human application. Because AAV vectors can persist in host cells without toxicity and integration into the genome, they are increasingly used for human gene therapy. To establish targeting of AAV vectors to liver myofibroblasts, we will screen a library of shuffled AAV capsids. In support of the feasibility of this approach, capsids targeting specifically pancreatic -cells have recently been identified in this library. By improvig function and reducing fibrosis of the liver with a viral vector deemed safe for human application, the proposed strategy has potential as an alternative to liver transplantation for therapy of liver failure due to alcoholic liver disease.